This project has been designed to integrate the efforts of a multi-disciplinary team of academic researchers from the Magnetic Resonance Science Center and the Brain Tumor Research Center at UCSF in order to translate a new metabolic imaging modality into a tool for clinical management of patients with gliomas. The objective of the study is to determine whether quantitative parameters derived from Magnetic Resonance Spectroscopic Imaging (MRSI) data are predictive of response to therapy for patients with gliomas. This is an important clinical question because gliomas are heterogeneous, infiltrative tumors with poorly defined margins. Although histological grade has been shown to be predictive of outcome in large-scale clinical trials, there is considerable variability between tumors of the same grade in terms of response to therapy and time to progression. While T2-weighted and Gadolinium enhanced Tl-weighted MR images are widely used for clinical evaluation of gliomas, they are only indirect measures of pathology and may both under- and over-estimate tumor burden. The identification of new factors that predict clinical outcome are critical for tailoring therapy to individual patient characteristics and are expected to have a significant impact upon the criteria used to select patients for future clinical trials. In Specific Aim 1, MRI and MRSI will be used to study the relationship between anatomic and metabolic lesions in a population of newly diagnosed patients with high-grade gliomas. The levels of choline (Cho), creatine (Cr) nd N-acetylaspartate (NAA), lactate (Lac) and lipid (Lip) have been shown to discriminate between tumor, normal brain and necrotic tissue and are expected to be helpful in determining tumor burden prior to surgery and in the postoperative period. Of particular interest is whether the metabolic indices that have been derived are correlated with response to therapy, time to tumor progression or survival. Clinically, this would be an important determinant in terms of stratification for future clinical protocols. In Specific Aim 2, we propose to evaluate the MRSI characteristics of gliomas for patients enrolled in prospective clinical therapeutic trials both before and after therapy. The goal will be to determine whether either the initial characteristics of the tumor or the changes following treatment are correlated with response to therapy. This would be a major advantage for tailoring therapies to individual patients, for sparing the potentially toxic effects of agents for patients who would not otherwise benefit and for understanding the mechanisms of success or failure of different therapies.